(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Biodiversity Heritage Library | Children's Library | Advanced Microdevices Manuals | Linear Circuits Manuals | Supertex Manuals | Sundry Manuals | Echelon Manuals | RCA Manuals | National Semiconductor Manuals | Hewlett Packard Manuals | Signetics Manuals | Fluke Manuals | Datel Manuals | Intersil Manuals | Zilog Manuals | Maxim Manuals | Dallas Semiconductor Manuals | Temperature Manuals | SGS Manuals | Quantum Electronics Manuals | STDBus Manuals | Texas Instruments Manuals | IBM Microsoft Manuals | Grammar Analysis | Harris Manuals | Arrow Manuals | Monolithic Memories Manuals | Intel Manuals | Fault Tolerance Manuals | Johns Hopkins University Commencement | PHOIBLE Online | International Rectifier Manuals | Rectifiers scrs Triacs Manuals | Standard Microsystems Manuals | Additional Collections | Control PID Fuzzy Logic Manuals | Densitron Manuals | Philips Manuals | The Andhra Pradesh Legislative Assembly Debates | Linear Technologies Manuals | Cermetek Manuals | Miscellaneous Manuals | Hitachi Manuals | The Video Box | Communication Manuals | Scenix Manuals | Motorola Manuals | Agilent Manuals
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Transcendence with Rosen continued fractions"

TRANSCENDENCE WITH ROSEN CONTINUED 

FRACTIONS 

YANN BUGEAUD, PASCAL HUBERT, AND THOMAS A. SCHMIDT 

o 

Cn ' Abstract. We give the first transcendence results for the Rosen con- 

. , tinned fractions. Introduced over half a century ago, these fractions 

^^Z ' expand real numbers in terms of certain algebraic numbers. 

m 









f— ( ■ 1. Introduction 

In 1954, D. Rosen defined an infinite family of continued fraction algo- 
rithms [2D]- Introduced to aid in the study of certain Fuchsian groups, these 
continued fractions were applied some thirty years later by J. Lehner [13] 
in the study of Diophantine approximation by orbits of these groups. 
K^ , The Rosen continued fractions and variants have been of recent inter- 

O ! est, leading to results especially about their dynamical and arithmetical 

(^ ! properties, see [9], [18], [H]; as well on their applications to the study 

. . of geodesies on related hyperbolic surfaces, see [23], [8], [17]; and to Te- 

(^ ■ ichm tiller geodesies arising from (Veech) translation surfaces, see [2B], [S], 

O' 



[25] and [6]. Several basic questions remain open, including that of arith- 
metically characterizing the real numbers having a finite Rosen continued 
fraction expansion, see [ISj, [13] and [B]. Background on Rosen continued 
^ ' fractions is given in the next section. 

The first transcendence criteria for regular continued fractions were proved 
by E. Maillet, H. Davenport and K. F. Roth, A. Baker, and recently im- 
proved by B. Adamczewski and Y. Bugeaud, see [H |2| and the references 



Date: 24 June 2010. 

2000 Mathematics Subject Classification. 11J70, 11J81 . 

Key words and phrases. Rosen continued fractions, Liouville inequality, Hecke groups, 
transcendence. 

The second named author is partially supported by project blanc ANR: ANR-06- 
BLAN-0038. The third author thanks FRUMAN, Marseille and the Universite P. 
Cezanne. 

1 



2 YANN BUGEAUD, PASCAL HUBERT, AND THOMAS A. SCHMIDT 

given there. In particular, Theorem 4.1 of [2] asserts that, if ^ is an alge- 
braic irrational number with sequence of convergents {pn/<in)n>i^ then the 
sequence {qn)n>i cannot increase too rapidly. It is natural to ask whether 
similar transcendence results can be proven using Rosen continued fractions. 
We give the first such results. 

Theorem 1.1. Fix A = 2cos7r/?7i for an integer m > 3 , and denote the 
field extension degree [Q(A) : Q] by D. If a real number C, ^ Q(A) has 
an infinite expansion in Rosen continued fraction over Q(A) of convergents 

Pn/qn satisfying 

log log g„ 
limsup > log(2L' — 1) , 

n— >oo ^ 

then E, is transcendental. 

For stating our second result, we associate to the Rosen continued fraction 
expansion 

[ei{x) : ri{x), e2{x) : r2(x), . . . , £„(x) : r„(x), . . .] 



£i 



riA 



^2 



rsA + ■ ■ ■ 

of a real number x the sequence of pairs of integers (^j, rj)j>i, which we call 
the partial quotients, and thus consider A = {±1} x N as the alphabet of 
the Rosen continued fraction expansions. 

As usual, we denote the length of a finite word U = ui ■ ■ ■ Uk a.s \U\ = k. 
For any positive integer s, we write [/* for the word U . . .U {s times repeated 
concatenation of the word U). More generally, for any positive real number 
s, we denote by U^ the word U^^^U', where U' is the prefix of U of length 
\is-[s\)\U\]. 

Just as Adamczewski and Bugeaud [1] showed for regular continued frac- 
tion expansions, a real number whose Rosen continued fraction expansion 
is appropriately "stammering" must be transcendental. 

Theorem 1.2. Fix A = 2cos7r/m- for an integer m > 3 , and denote the 
field extension degree [Q(A) : Q] by D. Let C, be an infinite Rosen continued 
fraction with convergents {Pn/qn)n>i such that 

B := lim sup g^" < +oo . 



TRANSCENDENCE WITH ROSEN CONTINUED FRACTIONS 3 

Write 

b := liminf g^" . 

n 

Assume that there are two infinite sequences {Un)n>i cind (\4,)n>i of finite 
words over the alphabet A and an infinite sequence {Wn)n>i of real numbers 
greater than 1 such that, forn > 1, the word UnV^^ is a prefix of the infinite 
word composed of the partial quotients of C,- If 

/,x ,. \Un\+Wn\Vn\ 3D \og B 
( 1 ) ^im sup ; ; ; > ■ , 

^ ^ n^+oo 2\Un\ + \Vn\ 2 log 6 ' 

then ^ is either (at most) quadratic over Q(A) or is transcendental. 
Lemma [2.11 gives that log 6 is positive. 

The key to our proofs is that both numerator and denominator of a 
Rosen convergent dominate their respective conjugates in an appropriate 
fashion, see Lemma 13.11 From this one can bound the height of a Rosen 
convergent in terms of its denominator, see Lemma 13.21 Then, exactly as 
in the case of regular continued fractions, we apply tools from Diophantine 
approximation, namely an extension to number fields of the Roth theorem, 



for the proof of Theorem II. H and the Schmidt Subspace Theorem for the 
proof of Theorem 11.21 

Both theorems are weaker than their analogues for regular continued frac- 
tions, since we have to work in a number field of degree D rather than in the 
field Q. However, for m = 4 and m = 6, that is, for A4 = ^/2 and Ag = v^, 
our results can be considerably strengthened and we can get, essentially, the 
exact analogues of the results established for the regular continued fractions. 
The key point is that, in both cases, for every convergent Pn/Qn, exactly one 
of Pn, Qn is in Z, the other being in XL , see Remark 2 below. 



2. Background 

2.1. Rosen fractions. We set A = Am = 2 cos— and Im = [—A/2, A/2) 
for m > 3. For a fixed integer m > 3, the Rosen continued fraction map is 



4 YANN BUGEAUD, PASCAL HUBERT, AND THOMAS A. SCHMIDT 

defined by 

T{x) = 

[o x = 

for X G Im! here and below, we omit tlie index "m" wlienever it is clear from 
context. For n > 1, we define 

en{x) = e(r"-V) and r„(x) = r{T''-\) 

with 

e{y) = sgn{y) and r{y) = 



Ay 



Then, as Rosen showed in [2U], the Rosen continued fraction expansion of 
X is given by 



>i(x) : ri(x), e2{x) : r2(x), . . . , en{x) : r„(x), ...]:= 



As usual we define the convergents Pn/Qn of x G I^ by 



riA + 



£2 



r2\ 



P^i Po 



1 
1 



and 



ei WO 62 

1 ArJ ll Xr2 



En 

1 Xrr^ 



Pn-l Pn 

for n > 1. From this definition it is immediate that IPn-iQn — Qn-iPnl = Ij 
and that the well-known recurrence relations 



p_i = 1; Po = 0; Pn = Ar„p„„i + enPn-2, n>l 
g_i = 0; go = 1; Qn = Ar„g„_i + €nqn-2, n > 1, 
hold. It also follows that 



(2) 

giving 

(3) 



Pn-l qn-1 
Pn Qn 



Pn 
Qn 



1 W 1 

^n ^fnj \£n-l Ar„_i 



[ei : n, 62 : r2, . . . , En : r„] 



1 

El Ari 



TRANSCENDENCE WITH ROSEN CONTINUED FRACTIONS 

and 

(4) ^^ = [ 1 : r„, £„ : r„_i, . . . , £2 : '^i] • 



Pn^l Vn 



We define 
(5) M„ = ( ^"^' ^" ) , and find that a; = M„ ■ T"(x) 

where ■ denotes the usual fractional linear operation, namely 

_ Pn-lT'^{x) +Pn 

g„_iT"(x) + Qn ' 

2.2. Approximation with Rosen fractions. We briefiy discuss the con- 
vergence of the "convergents" to x. One can rephrase some of Rosen's 
original arguments in terms of the (standard number theoretic) natural ex- 
tension map T(x, y) = {T{x), ^^]_^ ) where r = ri(x) and e = ei{x) . The 
"mirror formula" Equation (j4]) shows that T"'{x,y) = ( T" (x) , ^^^^^ ) . Ex- 
tending earlier work of H. Nakada, it is shown in [9] that T(x, y) has planar 
domain with y-coordinates between and R = -R(A) , where i? = 1 if the in- 
dex m is even and, otherwise, R is the positive root of i?^ + (2 — X)R — 1 = , 
in which case we have 1 > i? > A/2 (see Lemma 3.3 of [9]). Therefore, the 
sequence (g„)„>i is strictly increasing. But, as Rosen mentions, if x has 
infinite expansion, then either £„ = 1 or r„ > 1 occurs infinitely often; from 
this one has both that ^n > 1 for all n and that the limit as n tends to 
infinity of g„ is infinite. 

One easily adapts Rosen's arguments so as to find the following. 

Lemma 2.1. For every x G !,„ of infinite expansion, we have 

hm inf g^" > 1 . 

n 

Proof. We know that the sequence {qn)n>i increases and that, if either 
e„ = 1 or r„ > 1, then g„ > Ag„_i. Furthermore, there are no more than 
h consecutive indices i with {ei,ri) = (—1,1), with h = m/2,{m — 3)/2 
depending on the parity of m (see [20] or [5]). Consequently, for any n, 
there is some i = 1, . . . ,h + 1 such that qn+i > Ag„+j_i, giving 

Qn+h+l > Qn+i > ^Qn+i-l > Ag„ • 



YANN BUGEAUD, PASCAL HUBERT, AND THOMAS A. SCHMIDT 

1 



As gi > A, letting s{n 
this proves the lemma. 



1 + 



n 



h 



, we have g„ > A'^''"^ Since A > 1, 



n 



Remark 1. In fact, H. Nakada [19] shows that for almost all such x, 
liuin^r^ — log Qn exists, being equal to one half of the entropy of T. He also 



n 



shows that the entropy equals C ■ {m — 2)7r^/(2m), where C = 1/ log(l + R) 
when m is odd, and equals l/log[(l + cos7r/m)/sin7r/m] when m is even. 
This C is the normalizing constant to give a probability measure on the 
domain of the planar natural extension T, see [9]. 



Rosen also gave bounds on | x—pn/qn \ ■ Using Equation (^ (as in Nakada 



X 



Pn 

one has the easy lower bound 



„2 I 1n+l _|_ T^71+l™ 



Qn (g„,+l + Qn 

One also finds in this manner 

Pn 



< 



(6) 



Qn 



< 



X 



Cl 



Pn 

qn 



QnQn+1 



with Cl = Ci(A) = 2/(2 — RX) . Thus convergence does hold. 



Theorems 4.4 and 4.5 (depending on parity of m) of [9] give C2 such that 
I ^ " Pn/ln I < (^2/0^ ) with the upper bound C2 < 1/2 + [m/4] . 

2.3. Traces in Hecke groups. Rosen introduced his continued fractions 
to study the Hecke groups. The Hecke (triangle Fuchsian) group Gm with 
m G {3, 4, 5, . . . } is the group generated by 



1 A. 



and 



4 



^0 1 J \l 

with Am as above. The Rosen expansion of a real number terminates at a 
finite term if and only if a; is a parabolic fixed point of Gm , see [20]. These 
points are clearly contained in Q(Am) but in general there are elements of 
this field that have infinite Rosen expansion, see [15], [13] and [6]. 



TRANSCENDENCE WITH ROSEN CONTINUED FRACTIONS 7 

Remark 2. The values of finite Rosen expansion form the set Gq- oo, which 
is in fact a subset of A Q(A^) U {c)o}. To see this, one uses induction on word 
length in the generators displayed above — an ordered pair (a, c) giving a 
column of any element of Gq must be such that exactly one element of the 
pair is in Z[A^] , and the other is in AZ[A^] . Note that this also applies 
to convergents Pn/Qn'- exactly one oi Pn,qn is in ^[-^^l) the other being in 
AZ[A2] . 

When g = 3 , we have G3 = PSL(2, Z) . In general each Gm is isomorphic 
to the free product of a cyclic group of order two and a cyclic group of order 
m. Recall that a Fuchsian triangle group is generated by even words in the 
reflections about the sides of some hyperbolic triangle. Thus any Fuchsian 
triangle group is of index two in the group generated by these reflections; 
for each G^, we denote this larger group by A^ . 

Since \m is the sum of the root of unity ^2™ := exp27r2/(2m) with its 
complex conjugate, Q{Xm) is a number field of degree d := </)(2m)/2 over 
the rationals, where (p denotes the Euler totient function. 

The following key phenomenon property of Hecke groups can be shown 
in various manners. The result holds for a larger class of groups, from 
Corollary 5 of |2l] , due to [10] (extending the arguments from Gm to Am is 
straightforward). Independent of this earlier work, Bogomolny-Schmit [S] 
gave a clever proof of the result specifically for A^ . See the next remark 
for another perspective. 

Theorem 2.1. Fix m as above, and let Am. be the full reflection group in 
which Gm has index two. Then for any M G Am whose trace is of absolute 
value greater than 2, we have 

I tr{M) \>\(t{ tr{M) ) | , 

where a is any field embedding of Q{Xm) ■ 

Remark 3. This result can be proven "geometrically". Up to conjugacy, 
each of the Hecke groups appears as the Veech group of some translation 
surface, see [SB]; the elements whose trace is of absolute value at least 2 
are the "derivatives" of the affine pseudo-Anosov diffeomorphisms of the 



8 YANN BUGEAUD, PASCAL HUBERT, AND THOMAS A. SCHMIDT 

surface. The dilatation of a pseudo-Anosov is the dominant eigenvalue A 
of the action of on the integral homology of the underlying surface. (The 
other eigenvalues are hence conjugates of A.) The corresponding element of 
the Veech group has trace of absolute value A + A~^ from which it follows 
that this trace dominates its conjugates. 

2.4. Approximation by algebraic numbers. The following result was 
announced by Roth [22] and proven by LeVeque, see Chapter 4 of [16] . (The 
version below is Theorem 2.5 of [7].) Recall that given an algebraic number 
a, its naive height, denoted by H{a) , is the largest absolute value of the 
coefficients of its minimal polynomial over Z . 

Theorem 2.2. (LeVeque) Let K he a number field, and ^ a real algebraic 
number not in K . Then, for any e > , there exists a positive constant 
c(^, K, e) such that 



\^ - a\ > 



H(aY+' 



holds for every a in K. 



The logarithmic Weil height of a lying in a number field K of degree D 
over Q is h{a) = jj ^^ log^ max,^gj\/^^,{| |a| |,^}, where log"*" t equals if t < 1 
and Mk denotes the places (finite and infinite "primes") of the field, and 
II ■ III, is the //-absolute value. This definition is independent of the field K 
containing a. The two heights are related by 

(7) log H{a) < deg{a)h{a) + log 2, 

for any non-zero algebraic number a, see Lemma 3.11 from [27] . 

We recall a consequence of the W. Schmidt Subspace Theorem. 

Theorem 2.3. Let d be a positive integer and ^ be a real algebraic number 
of degree greater than d. Then, for every positive e, there exist only finitely 
many algebraic numbers a of degree at most d such that 



1^ - a| < H{a) 



-d-l-e 



TRANSCENDENCE WITH ROSEN CONTINUED FRACTIONS 

Note that the Roth theorem is exactly the case d = 1 oi Theorem 



In the proof of Theorem 11.21 we could apply Theorem 12.31 but the al- 
gebraic numbers a which we use to approximate ^ are of degree at most 2 
over a fixed number field. In this situation, the next theorem, kindly com- 
municated to us by Evertse [12], yields a stronger result than the previous 
one. 

Theorem 2.4. (Evertse) Let K he a real algebraic number field of degree d. 
Let t be a positive integer and ^ be a real algebraic number of degree greater 
than t over K . Then, for every positive e, there exist only finitely many 
algebraic numbers a of degree t over K and 5 over Q such that 

Note that Theorem 12.41 extends Theorem 12.21 

2.5. Sturmian sequences: towards an application of Theorem 1.2. 

To give an explicit family of Rosen expansions satisfying the hypotheses of 
Theorem 1.2, we recall a result of |3] on Sturmian sequences. 

Let a and h be letters in some alphabet. The complexity function of a 
sequence u = U1U2 ■ ■ ■ with values in {a, h} is given by letting p{n, u) be the 
number of distinct words of length n that occur in u. A sequence u is called 
Sturmian if its complexity satisfies p(n, u) = n + 1 for all n. As Arnoux 
[1] writes, one can obtain any such sequence by taking a ray with irrational 
slope in the real plane and intersecting it with an integral grid, assigning a 
when the ray intersects a horizontal grid line and h when it meets a vertical 
grid line. Indeed, the slope of a Sturmian sequence is the density of a in the 
sequence (one shows that the limit as n tends to infinity of the average of 
the number of occurrences a'mui- ■ -Un exists, see |1], Proposition 6.1.10). 

Lemma 2.2. Let u be a Sturmian word whose slope has an unbounded 
regular continued fraction expansion. Then, for every positive integer n, 
there are finite words U, V and a positive real number s such that UV^ is 
a prefix of u and \ UV^ | > '^ | UV \ . 

Proof. This follows from the proof of Proposition 11.1 from [3]. D 



10 YANN BUGEAUD, PASCAL HUBERT, AND THOMAS A. SCHMIDT 

Remark 4. We apply the above lemma to Sturmian sequences where both 
a, b are of the form {e, r), with e = ±1 and r G N. In particular, we use this 
in the context of Rosen expansions to prove Corollary 14.11 

3. Bounding the height of convergents 

In what follows, we fix A = A^ for some ?7i > 3, and suppose that 
^ G (0, A/2) is a real algebraic number having an infinite Rosen continued 
fraction expansion over Q(A) . Our goal is to estimate the naive height 
H{pn/qn) of the nth convergent Pn/Qn- In light of Theorem l2.lt we let no 
be the least value of n such that qn > 2 . 

Lemma 3.1. Let c^ = c^iX) be defined by c^ = miug- \^ , where the 
minimum is taken over all field embeddings of Q(A) into M . Then for all 
n > hq, and any such a , we have both 

gn > C3 I a{qn) I and p„ > C3 | a{pn) \ ■ 

Proof. For any n > uq, recall that M„ = ( "~ "■ ) ; this is clearly an 

element of A„i . By Theorem 12.11 we have g„ + Pn-i > | (^{qn + Pn-i) I • 
Now let j G N and set 

%,_1 Pn\ [l jX\ fpn-l Pn+jXpn-l 



"'^ \qn-l qnj \0 I J \qn-l qn+jMn-l 

This is also an element of A^ of trace greater than 2, and hence 

I Pn~i + qn+ jAg„_i I > I cr(p„,„i + g„) + jcr(Ag„_i) I . 

Since this holds for all positive j , we must have that Ag„_i > | a{\qn-i] 
That is, 

qn^i > — T — \cr{qn^i)\ > I mm — - — I |fT(g„_i)| . 

Similarly, using 



we find 



Pn^l Pn\ Z' 1 0\ fPn^l+jXPn Pn 

qn^i qn) XJX 1/ \qn^i+3Mn qn 



Pn > — T — WiPnJl > I mm — - — I \a{pn)\ 



D 



TRANSCENDENCE WITH ROSEN CONTINUED FRACTIONS 11 

Remark 5. We conjecture that in fact g„ is always greater than or equal 
to its conjugates, thus that in the above one can replace C3 by 1. 

Lemma 3.2. Let D denote the field extension degree [Q( A ) : Q] . There 
exists a constant C4 = C4(A) such that for all n > Uq, 

Proof. Since Pn and g„ are algebraic integers of degree at most D, it follows 
from Lemma [3. II that 

h{pn/qn) ^^ Jj logmax{|a(|)„)|, \a{qn)\} <c[ + logg„, 

(T 

where a runs through the complex embeddings, for a suitable positive con- 
stant C4. Using d?]), we get the asserted estimate. D 

Lemma 3.3. Let a be a real number in [—A/2, A/2) with an ultimately 
periodic expansion in Rosen continued fraction. Denote by {pn/<ln)n>i the 
sequence of its convergents. Denote by fi the length of the preperiod and by 
V the length of the period, with the convention that fi = if the expansion is 
purely periodic. Then a is of degree at most 2 over Q(A), and there exists 
C5 = C5(A,a) such that 

H{a) < C5(g^g^+^)^. 

Proof. In the notation of Equation (E]), a is fixed by M = M~^M^+j,. It thus 
satisfies a quadratic equation with entries in Z[A], and hence is of degree 
at most 2 over Q(A). Indeed, a is a root of /(x) = ex + {d — a)x — b with 
a, b, c, d denoting the entries of M. Each entry is a Z-linear combination of 
monomials of the form rs with r an entry of M^ and s an entry of M^j^y. 

Now, a is also a root of /(x) = Y\^ a{f){x) G Z[x], where a{f) denotes 
the result of applying a to the coefficients of f{x). By Lemma |3TT| all of the 
conjugates of each of p^,p^_i, g^_i, q^ can be bounded by the product of q^ 
with a constant depending upon a and A. Similarly for the entries of M^j^^. 
After some computation, we conclude that the height of a is <^ QuQu+u- 
(One checks that the case of /i = is subsumed by the above.) D 

Remark 6. Whereas a real number whose regular continued fraction expan- 
sion is ultimately periodic is exactly of degree two over the field of rational 



12 YANN BUGEAUD, PASCAL HUBERT, AND THOMAS A. SCHMIDT 

numbers, in the previous lemma the words "at most" are necessary. Indeed, 
X = 1 has an ultimately periodic Rosen expansion with respect to any A^ 
with m even, [2D]. Further examples of elements of Q(Am) with periodic 
expansions are easily given when m G {4,6}, see Corollary 1 of [23]. Yet 
further examples, including cases with m G {7,9}, are given in [21], [13] . 

4. Transcendence results 

As usual, <^ and ^ denote inequality with implied constant. 

4.1. Applying Roth— LeVeque: the proof of Theorem 11.11 We now 
show that the sequence of denominators of convergents to an algebraic num- 
ber cannot grow too quickly. Theorem 11.11 then follows. 

Proof. Let £ be a positive real number. Let ( be an algebraic number having 
an infinite Rosen expansion with convergents Vn/sn- 
By the Roth-LeVeque Theorem 12. 2[ we have 

|C-^n/sn| >^(r„/s„)~2-^ forn>l. 

And, hence by Lemma 13. 2^ for n > uq = uq^O, we have \( — r.„/s„| 3> 
s'"^^'^^ . Inequality (J6]) then gives that there exists a constant cq (indepen- 
dent oi n > uq) such that 

„ ^ „ „2D-1+De 

Sn+1 < CgS^ 

Set a = 2D — 1 + De . For j < uq, define ij such that Sj < (-jS°j_i . We 
set Ci = max{l, cq, ii, . . . , ^no-i} and find that for any n > 1 

and continuing in this manner, we have s„+i < (cySi)"". Since s„+i > s„ 

letting Cg = cySi, gives logs„ < a^logcg . From this follows that 

loglogSn 
limsup < log(Z;(2 + e) — 1). 

n— >-+oo IT, 

Letting e go to zero, we see that every algebraic number satisfies 

log log s„ 
limsup < log(2i? — 1), 

n— >+oo IT, 

as asserted. D 



TRANSCENDENCE WITH ROSEN CONTINUED FRACTIONS 13 

4.2. Proof of Theorem 11.21 and an application. 

Proof. With A = 2 cos n/m fixed, given ^ of infinite Rosen continued fraction 
with convergents {pn/c[n)n>i, we let b = liminf„g„ and B = limsup^gn , 
and assume that B < oo. Let r/ be a positive real number with 6—1 < rj < b. 
Since there are only finitely many n with either g„ < b — rfoi qn > B + r], 
we have both that qn^ ip — r])"' and g„ ^ (5 + 77)'" . 

Suppose that u; is a positive real number and U, V are finite words in 
{±1} X N such that UV"^ is a prefix of the infinite word composed of the 
partial quotients of ^. Denote by a the real number of degree at most two 
over Q(A) whose Rosen continued fraction is given by the word UV°°, where 
V^ means the concatenation of infinitely many copies of V. Set \U\ = u 
and 1 1^1 = V. Since ^ and a have their first [u + vw\ partial quotients in 
common, we have 

Furthermore, it follows from Lemma [3.31 that 

Hia) « {qu qu+.)'' « {B + r/)^(2.+.)^ 
Combined with the previous inequality, this gives 

1^ - al < J^l^\-2{u+vw)\og{b-r))/{D{2u+v)\og{B+ri))^ 

Now suppose that ^ is algebraic of degree greater than two over Q(A). 
Then, for every e > 0, there exists a positive constant C{e) such that every 
real algebraic number /3 of degree at most 2 over Q(A) satisfies 



\i-P\>C{e)H{P) 



-3-e 



This follows from Theorem 12.21 if /3 is in Q{\) and, otherwise, by applying 
Theorem [Ml with t = 2 and t/t = 5 to each subfield K of Q(A). 

This proves that ^ must be transcendental if there are m, v, w such that 
u + vw is arbitrarily large and 

2{u + vw)\ogb 
D{2u + v)\ogB ^ ' 

as asserted. D 



14 YANN BUGEAUD, PASCAL HUBERT, AND THOMAS A. SCHMIDT 

Corollary 4.1. A Rosen continued fraction whose sequence of partial quo- 
tients is Sturmian with slope of unbounded regular continued fraction partial 
quotients represents a transcendental number. 

Proof. Combine Lemma 12.21 with the previous Theorem. D 

Remark 7. Using the Subspace Theorem as in [1] does not yield in general 
an improvement of Theorem II. 2[ In case u = 0,b = B, inequality (II]) 
reduces to w > 3D/2, while, proceeding as in [1], we would get w > 2D — 1. 
However, if b is much smaller than B and D is small, then the approach of 
[l] presumably gives a slightly better result than Theorem 11.21 



References 

[1] B. Adamczcwski and Y. Bugcaud, On the complexity of algebraic numbers II. Con- 
tinued fractions, Acta Math. 195 (2005), 1-20. 

[2] , On the Maillet-Baker continued fractions, J. Reine Angew. Math. 606 

(2007), 105-121. 

[3] , Nombres reels de complexite sous-lineaire : mesures d'irrationalite et de 

transcendance. J. Rcine Angcw. Math., to appear. 

[4] P. Arnoux, Sturmian sequences, in "Substitutions in dynamics, arithmetics and com- 
binatorics", 143-198, Lecture Notes in Math., 1794, Springer, Berhn, 2002. 

[5] P. Arnoux and P. Hubert, Fractions continues sur les surfaces de Veech, J. AnaL 
Math. 81 (2000), 35-64. 

[6] P. Arnoux and T. A. Schmidt, Veech surfaces with non-periodic directions in the 
trace field, J. Mod. Dyn. 3 (2009), no. 4, 611-629. 

[7] Y. Bugeaud, Approximation by algebraic numbers, Cambridge Tracts in Mathemat- 
ics, 160. Cambridge University Press, Cambridge, 2004. 

[8] E. Bogomolny and C. Schmit, Multiplicities of periodic orbit lengths for non- 
arithmetic models, J. Phys. A: Math. Gen. 37, (2004) 4501-4526. 

[9] R. Burton, C. Kraaikamp, and T.A. Schmidt, Natural extensions for the Rosen 
fractions, TAMS 352 (1999), 1277-1298. 
[10] P. Cohen and J. Wolfart, Modular embeddings for some nonarithmetic Fuchsian 

groups. Acta Arith. 56 (1990), no. 2, 93-110. 
[11] K. Dajani, C. Kraaikamp, W. Steiner, Metrical theory for a-Rosen fractions, J. Eur. 

Math. Soc. (JEMS) 11 (2009), no. 6, 1259-1283. 
[12] J.-H. Evertse, private communication. 

[13] E. Hanson, A. Merberg, C. Towse, and E. Yudovina, Generalized continued fractions 
and orbits under the action of Hecke triangle groups. Acta Arith. 134 (2008), no. 4, 
337-348. 
[14] J. Lehner, Diophantine approximation on Hecke groups, Glasgow Math. J. 27 (1985), 

117-127. 
[15] A. Leutbecher, Uber die Heckeschen Gruppen G(X), Abh. Math. Sem. Hamb. 31 
(1967), 199-205. 



[16; 

[17; 
[is; 

[19 

[2o; 

[21 
[22; 
[23; 
[24 
[25; 

[26; 
[27; 



TRANSCENDENCE WITH ROSEN CONTINUED FRACTIONS 15 

W. J. LeVeque, Topics in number theory, Vols. 1 and 2. Addison- Wesley Publishing 

Co., Inc., Reading, Mass., 1956. 

D. Mayer and F. Stromberg, Symbolic dynamics for the geodesic flow on Hecke 

surfaces, J. Mod. Dyn. 2 (2008), no. 4, 581-627. 

H. Nakada, Continued fractions, geodesic flows and Ford circles, in Algorithms, 

Fractals and Dynamics edited by Y. Takahashi, 179-191, Plenum, 1995. 

, On the Lenstra constant associated to the Rosen continued fractions, J. Eur. 

Math. Soc. (JEMS) 12 (2010), no. 1, 55-70. 

D. Rosen, A class of continued fractions associated with certain properly discontin- 
uous groups, Duke Math. J. 21 (1954), 549-563. 

D. Rosen, C. Towsc, Continued fraction representations of units associated with 
certain Hecke groups. Arch. Math. (Basel) 77 (2001), no. 4, 294-302. 
K. F. Roth, Rational approximations to algebraic numbers, Mathematika 2 (1955) 
1-20; corrigendum 168. 

T. A. Schmidt, M. Shcingorn, Length spectra of the Hecke triangle groups, Math. Z. 
220 (1995), no. 3, 369-397. 

P. Schmutz Schaller and J. Wolfart, Semi- arithmetic Fuchsian groups and modular 
embeddings, J. London Math. Soc. (2) 61 (2000), no. 1, 13-24. 

J. Smillie and C. Ulcigrai, Geodesic flow on the Teichmiiller disk of the regular 
octagon, cutting sequences and octagon continued fractions maps, preprint (2010): 
■arXiv:1004.2265' 

W.A. Vecch, Teichmiiller curves in modular space, Eisenstein series, and an appli- 
cation to triangular billiards, Inv. Math. 97 (1989), 553 - 583. 

M. Waldschmidt, Diophantine Approximation on Linear Algebraic Groups. Tran- 
scendence properties of the exponential function in several variables. Grundlehren 
der Mathematischen Wissenschaften 326. Springcr-Verlag, Berlin, 2000. 



Universite de Strasbourg, Mathematiques, 7, rue Rene Descartes, 67084 
Strasbourg cedex, France 

E-mail address: bugeaud@math.unistra.fr 

LATP, CASE couR A, Faculte des sciences Saint Jerome, Avenue Escadrille 

NORMANDIE NiEMEN, 13397 MARSEILLE CEDEX 20, FRANCE 

E-mail address: liubert@cini.urLiv-mrs.fr 

Department of Mathematics, Oregon State University, Corvallis, OR 
97331, USA 

E-mail address: toms@math.orst.edu